Results of Cox Proportional Hazards Model for Covariate Analysis of Censored Data on Survival (P values)
. | All Patients (n = 306) . | IFN-α Patients (n = 162) . | ||
---|---|---|---|---|
Univariate . | Multivariate . | Univariate . | Multivariate . | |
Sex | .332 | .558 | .214 | .255 |
Age | .002 | .003 | .025 | .098 |
Spleen | .002 | .006 | .002 | .344 |
Hemoglobin | .009 | .435 | .052 | .857 |
Platelet | .02 | .153 | .055 | .568 |
WBC | .005 | .624 | .051 | .406 |
Peripheral blood blast cells (%) | .003 | .348 | .024 | .034 |
Sokal’s score | .0001 | .001 | .0001 | .0001 |
Treatment arm | .001 | .001 | — | — |
Hematologic response4-150 | — | — | .0001 | .002 |
Cytogenetic response4-151 | — | — | .001 | .001 |
. | All Patients (n = 306) . | IFN-α Patients (n = 162) . | ||
---|---|---|---|---|
Univariate . | Multivariate . | Univariate . | Multivariate . | |
Sex | .332 | .558 | .214 | .255 |
Age | .002 | .003 | .025 | .098 |
Spleen | .002 | .006 | .002 | .344 |
Hemoglobin | .009 | .435 | .052 | .857 |
Platelet | .02 | .153 | .055 | .568 |
WBC | .005 | .624 | .051 | .406 |
Peripheral blood blast cells (%) | .003 | .348 | .024 | .034 |
Sokal’s score | .0001 | .001 | .0001 | .0001 |
Treatment arm | .001 | .001 | — | — |
Hematologic response4-150 | — | — | .0001 | .002 |
Cytogenetic response4-151 | — | — | .001 | .001 |
The model was applied to all cases, with the exclusion of 16 patients who were transplanted or were shifted to other treatments within the first 8 months. For IFN-α arm, the model was applied to the 162 patients (74% of total) who were alive and in chronic phase after 24 months, to avoid the bias related with time-dependent variables like hematologic and cytogenetic response (landmark analysis at 24 months).
Hematologic response was evaluated after 8 months.
Cytogenetic response was defined as a percentage of Ph neg metaphases ≥33% and was evaluated within 24 months, with examinations at 8, 14, and 24 months.